INDEX

Explanations

No Explanations Found

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 0-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.hook_resid_pre

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

24,576

Data Type

torch.float32

Hook Point

blocks.0.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

ahi

-0.75

nah

-0.74

Cu

-0.73

chrome

-0.73

iam

-0.69

ize

-0.69

ja

-0.68

aila

-0.68

lee

-0.67

jar

-0.67

POSITIVE LOGITS

 some

0.84

 SOME

0.74

avorite

0.70

 inconven

0.70

 unintention

0.69

Þ

0.66

exh

0.66

 fict

0.66

 stumbling

0.66

ONSORED

0.65

Activations Density 0.000%

No Known Activations

This feature has no known activations.